M E TAB O LI S M CL IN I CA L A N D EX PE RI ME N TA L 6 5 ( 2 01 6 ) 5 0 7–5 21

Available online at www.sciencedirect.com

Metabolism
www.metabolismjournal.com

The evolution of diabetic ketoacidosis: An update of

its etiology, pathogenesis and management

Ebenezer A. Nyenwe⁎, Abbas E. Kitabchi

Division of Endocrinology, Diabetes and Metabolism, University of Tennessee Health Science Center, 920 Madison Ave., Suite 300A, Memphis,
TN 38163

A R T I C LE I N FO AB S T R A C T

Article history: The prognosis of diabetic ketoacidosis has undergone incredibly remarkable evolution since
Received 6 October 2015 the discovery of insulin nearly a century ago. The incidence and economic burden of
Accepted 16 December 2015 diabetic ketoacidosis have continued to rise but its mortality has decreased to less than 1%
in good centers. Improved outcome is attributable to a better understanding of the
Keywords: pathophysiology of the disease and widespread application of treatment guidelines. In this
Diabetic ketoacidosis review, we present the changes that have occurred over the years, highlighting the evidence
Etiology behind the recommendations that have improved outcome. We begin with a discussion of
Pathogenesis the precipitants and pathogenesis of DKA as a prelude to understanding the rationale for
Management the recommendations. A brief review of ketosis-prone type 2 diabetes, an update relating to
the diagnosis of DKA and a future perspective are also provided.
© 2016 Elsevier Inc. All rights reserved.

1. Background DKA was invariably fatal until the discovery of insulin in

Diabetic Ketoacidosis (DKA) is a potentially fatal metabolic
complication of uncontrolled diabetes mellitus. In his first
clinical description of diabetes mellitus in the 2nd Century
A.D. Aretaeus gave a detailed account of subjects with
hyperglycemic crises [1], but it was Julius Dreschfeld, a
German pathologist who further characterized DKA in his
lecture to the Royal College of Physicians in London in 1886.
He reported on the main ketones, acetoacetate and β-
hydroxybutyrate, and their chemical determination [2]. The
incidence of DKA in developed countries is comparable with
estimated annual incidence rate of 13.6 and 14.9 per 1000 type
1 diabetic patients in the UK [3] and Sweden [4] respectively
and 13.4 per 1000 subjects younger than 30 years in the US [5].
Hospital admission for DKA has increased by about 75% over
the last two decades in the USA from about 80,000 in 1988 to
140,000 in 2009 [6].
the 1920s; however, DKA related mortality has reduced
significantly over the years. In the US, the age-adjusted
mortality rate decreased 64% from 48.4 per 100,000 diabetic
population in 1980 to 17.3 per 100,000 diabetic population in
2009 [6]. Mortality was also reported to be low in the Europe
with one UK institution recording no deaths amongst 46 DKA
patients between 1997 and 1999 [3]. Overall, the mortality in
adults in the UK and USA is less than 1% [3,6], but may be
higher than 5% in the elderly and patients with severe co-
morbid conditions [7,8]. DKA remains a leading cause of
mortality in children and young adults with type 1 diabetes
[9,10]. Morbidity and mortality from DKA remain high in
developing countries, with incidence of about 80 per 1000
diabetic admissions and mortality rate of 30% in Kenya [11]
and incidence of 41.7 per 100,000 population and mortality
rate of 11.7% in Libya [12]. DKA is economically burdensome
with an average length of stay of 3.4 days, DKA is responsible

⁎ Corresponding author at: Division of Endocrinology, Diabetes and Metabolism, University of Tennessee Health Science Center, 920
Madison Ave., Suite 300A, Memphis, TN 38163. Tel.: +1 901 448 2610; fax: + 1 901 448 4340.
E-mail addresses: enyenwe@uthsc.edu, eanyenwe@yahoo.com (E.A. Nyenwe).

http://dx.doi.org/10.1016/j.metabol.2015.12.007
0026-0495/© 2016 Elsevier Inc. All rights reserved.
508 M ET AB O LI S M CL IN I CA L A N D E XP E RI ME N TAL 6 5 ( 2 01 6 ) 5 0 7–52 1

Table 1 – Diagnostic criteria and typical total body deficits conditions contributes to better outcome. Mortality in DKA is
of water and electrolytes in diabetic ketoacidosis. usually related to the associated co-morbidity rather than the
Mild Moderate Severe biochemical derangement. Omission or inadequate dosing of
insulin and infection are the most common precipitants of
Diagnostic criteria and
DKA [19]. More recent reports may suggest that omission of
classification
insulin, which is preventable, is becoming a more frequent
Plasma glucose (mg/dl) + >250 >250 >250 precipitant of DKA than infections. Intercurrent illness such
Arterial pH 7.25–7.30 7.00- < 7.24 <7.00 as cerebrovascular accident, pancreatitis, myocardial infarc-
Serum bicarbonate 15–18 10- <15 <10 tion, trauma, and drugs are well known to trigger DKA. Drugs
(mEq/L) that affect carbohydrate metabolism such as corticosteroids,
Urine ketone ⁎ Positive Positive Positive
thiazides, and sympathomimetic agents like dobutamine and
Serum ketone ⁎ Positive Positive Positive
Effective Serum Variable Variable Variable
terbutaline as well as atypical antipsychotic agents could
Osmolality ⁎⁎ precipitate DKA in susceptible individuals [20]. Subjects with
Anion Gap ⁎⁎⁎ >10 >12 >12 type 1 diabetes using amphetamine-like analogs may be
Mental Status Alert Alert/Drowsy Stupor/Coma predisposed to DKA due to elevated catecholamine levels [21].
Emerging data suggests Sodium-Glucose Cotransporter 2
Typical deficits
Inhibition may increase the risk of DKA, prompting the FDA
Total Water (L) 6
to issue a warning in this regard in May 2015 [16,22]. Twenty
Water (ml/kg) δ 100
Na + (mEq/kg) 7–10 cases of DKA were reported with SGLT-2 inhibitors in patients
Cl- (mEq/kg) 3–5 with diabetes in the FDA Adverse Event Reporting System
K+ (mEq/kg) 3–5 (FAERS) between March 2013 and June 6, 2014. The mecha-
PO4 (mmol/kg) 5–7 nism of DKA in subjects treated with SGLT2 is not known with
Mg ++ (mEq/kg) 1–2 certainty; putative mechanisms include reduced insulin dose,
Ca ++ (mEq/kg) 1–2
glucagon secretion and decreased excretion of ketone bodies
+
Euglycemic DKA has been reported. [22]. In young patients with type I diabetes, psychological
⁎ Nitroprusside reaction method. problems complicated by eating disorders may be a contrib-
⁎⁎ Calculation: Effective serum osmolality: 2[measured Na +
uting factor in 20% of recurrent ketoacidosis. Other factors
(mEq/L) + glucose (mg/dl)/18 [mOsm/kg].
that may lead to insulin omission and DKA in younger
⁎⁎⁎ Calculation: Anion Gap: (Na+)–(Cl− + HCO−3 (mEq/L) [normal =12 ± 2].
δ patients include fear of weight gain and hypoglycemia,
Per kg of body weight. Data adapted from ref [19].
rebellion against authority and the stress of chronic disease
[23]. Cocaine use was reported as an independent risk factor
for recurrent DKA in a retrospective study of over 200 cases of
for about half a million hospital days per year and an estimated DKA [24]. Another study of a large cohort of DKA patients
annual direct and indirect cost of 2.4 billion USD [6,13]. observed that substance abuse especially cocaine but also
alcohol and cannabis were associated with recurrent episodes
DKA consists of the biochemical triad of hyperglycemia,
of DKA [25]. A recent report [26], suggested a relationship
ketonemia and metabolic acidosis (table 1) resulting from
between low carbohydrate dietary intake and metabolic acidosis.
absolute or relative insulin deficiency in the presence of an
increase in counterregulatory hormones (glucagon, catechol- Additionally, mechanical problems with continuous subcutane-
amines, cortisol, and growth hormone). Although DKA is ous insulin infusion devices (CSII) has also been associated with
typically characterized by hyperglycemia, euglycemic DKA has DKA [27]. Finally, DKA has also been reported as the initial
been reported in patients with type 1 diabetes who were manifestation of previously undiagnosed endocrine conditions
vomiting, fasting or had been treated with insulin prior to like acromegaly [28] and pheochromocytoma [29,30].
presentation, and in pregnancy [14,15]. Additionally, there has
been recent reports of euglycemic DKA in subjects treated with
sodium–glucose cotransporter 2 (SGLT2) inhibitors [16,17]. 3. Ketosis-Prone Type 2 Diabetes
DKA is more common in subjects with type 1 diabetes; but can
also occur in type 2 diabetes, especially in patients of African or African authors reported about temporary diabetes in adults
Hispanic descent [18]. About 35% of DKA cases in the USA in 2006 in the 1960s, subjects who after an episode of DKA could
occurred in people with type 2 diabetes [6]. Similarly, a study from maintain glycemic control for varying periods without insulin
Sweden noted that type 2 diabetes accounted for 32% of 26 therapy [31,32]. More recently, an increasing number of DKA
episodes of DKA recorded in that Caucasian population, further- cases with no apparent precipitating factors have been
more, in 50% of patients with type 2 diabetes, DKA was the initial reported in subjects with type 2 diabetes; studies have
manifestation of diabetes [4]. indicated that about half of previously undiagnosed adult
African Americans (AAs) and Hispanic subjects with unpro-
voked DKA have type 2 diabetes [33–37]. The evidence for type
2. Precipitating Factors 2 diabetes in these patients include obesity, family history of
diabetes, relatively preserved insulin secretion, low preva-
A diligent investigation for the precipitating illness should be lence of beta cell autoimmunity, and the ability to discontinue
made in all cases of DKA, as effective treatment of these insulin therapy during the period of near-normoglycemia [38–40].
 M E TAB O LI S M CL IN I CA L A N D EX PE RI ME N TA L 6 5 ( 2 01 6 ) 5 0 7–5 21
This variant of type 2 diabetes has been referred to as
idiopathic type 1 diabetes, atypical diabetes mellitus,
Flatbush diabetes, type 1.5 diabetes, and more recently as
ketosis-prone type 2 diabetes [18,41]. Ketosis-prone type 2
diabetes was thought to be present in AA alone but has been
reported in Caucasians, Hispanics, Chinese, South Asians,
and sub-Saharan Africans; AAs and Hispanics have a higher
risk than Caucasians [40]. During DKA, subjects with ketosis-
prone type 2 diabetes exhibit profound impairment in insulin
secretion and action, but they recover beta-cell function and
insulin sensitivity after resolution of DKA. Thus they are able
to discontinuing insulin therapy within a few months of
achieving near-normoglycemia [18,34,35,40]. Predictors of
remission include absence of autoimmunity with preserved
beta cell function as determined by a fasting C-peptide level
of > 1.0 ng/dl (0.33 nmol/l) or a glucagon-stimulated C-pep-
tide of > 1.5 ng/dl (0.5 nmol/l) [34,40]. Remission is associated
with a greater recovery of basal and stimulated insulin
secretion; ten years after onset of diabetes, 40% of patients
with ketosis-prone type 2 diabetes are still able to maintain
glycemic control without insulin therapy [35].
The pathophysiologic reason for acute but transient ß-cell
failure is not known with certainty. Postulated mechanisms
include glucotoxicity, lipotoxicity and genetic predisposition.
Sustained hyperglycemia is known to impair beta cell
function and elevated free fatty acid level has been implicated
in insulin resistance and inappropriate beta cell response
[42,43]. However, subjects with ketosis-prone type 2 diabetes
did not show any impairment in stimulated insulin secretion
after 20 h of hyperglycemia achieved by glucose infusion [44].
Furthermore, intralipid infusion for 48 h, which increased
blood free fatty acid level fourfold did not produce acute beta
cell failure [45]. Diminished insulin action in ketosis-prone
type 2 diabetes may be mediated through AKT-2. Insulin
stimulated phosphorylation and protein expression was
found to be reduced in muscle of subjects with ketosis prone
type 2 diabetes, defects that improved at the time of
remission of hyperglycemia [46]. Immunogenetic studies
appear concordant on the lack of autoantibodies preponder-
ance in subjects with ketosis-prone type 2 diabetes [33,38,48];
but reports are conflicting on HLA association. While two
studies [33,38] found no association, another study reported
association with HLA-DR3 and HLA-DR4 [48]. A few studies
have investigated putative candidate genes that may be
associated with ketosis-prone type 2 diabetes. A point
mutation in the HNF-1 gene was suspected to be marker in
AAs [49], but this finding was not replicated in adults of sub-
Saharan and Afro-Caribbean descent who instead had high
frequency of polymorphism in the transcription factor PAX4
[50]. Furthermore, a higher prevalence of glucose-6 phospha-
tase deficiency has been reported in subjects with ketosis-
prone diabetes and there was a proportional relationship
between β-cell functional reserve and erythrocyte glucose-6
phosphatase activity but increased prevalence of glucose-6
phosphatase mutation was not found in these patients [51].
Majority of patients with ketosis-prone diabetes achieve
remission but become increasingly insulin dependent over time
if treated with lifestyle modification alone. Recurrence of ketosis
occurs within 12–24 months in nearly 60% of patients who are
treated with lifestyle modification alone [36,40,47]. Small dose
509
sulfonylureas such as glyburide 1.25 to 2.5 mg/day) [47] and
glipizide 2.5 mg/day [48] and the thiazolidinedione pioglitazone
30 mg/day [40] have been shown to prolong remission and
prevent recurrence of DKA.
4. Pathogenesis
DKA results in abnormal metabolism of carbohydrate, pro-
tein, fat and derangement of fluid and electrolyte homeosta-
sis. The fundamental pathogenetic mechanism is a decrease
in the net effective action of circulating insulin, in the
presence of elevated counter-regulatory stress hormones
such as glucagon, epinephrine, norepinephrine, cortisol, and
growth hormone. Elevated glucagon level plays a major role in
the pathogenesis of DKA but it is not indispensable in the
development of this condition as totally pancreatectomized
subjects developed DKA when deprived of insulin [52].
4.1. Carbohydrate Metabolism
In the presence of decreased net effective action of insulin,
hyperglycemia results from increased gluconeogenesis, glyco-
genolysis and impaired peripheral glucose utilization due to
insulin resistance [53,54]. The gluconeogenic enzymes fructose
1,6 biphosphatase, phosphoenolpyruvate carboxykinase (PEPCK),
glucose 6 phosphatase and pyruvate carboxylase, are stimulated
by increased glucagon:insulin ratio and hypercortisolism resulting
in accelerated hepatic glucose production [55] (see Fig. 1). There is
also an increase in gluconeogenic precursors such as the
aminoacids alanine and glutamine, as a result of protein
catabolism; lactate from increased muscle glycogenolysis and
glycerol from increased lipolysis [53]. Hepatic gluconeogenesis is
the main mechanism for hyperglycemia in ketoacidosis, howev-
er, recent studies indicate renal gluconeogenesis may be a
contributing factor [56]. Hyperglycemia may be mild in subjects
who maintain good renal function, which supports glycosuria;
but as the disease process evolves, glucose-induced osmotic
diuresis leads to volume depletion and a reduction in glomerular
filtration thus impeding further glucose excretion [57].
4.2. Lipid and Ketone Metabolism
Reduced effective insulin action and increased concentrations of
counterregulatory hormones, especially epinephrine, which acti-
vates hormone-sensitive lipase in adipose tissue leads to the
increased production of non-esterified fatty acids (NEFA) and
glycerol from breakdown of triglycerides in DKA [58]. Glycerol is
used as a substrate for gluconeogenesis but the release of NEFA
assumes pathophysiological prominence in the liver. NEFA are
oxidized to ketone bodies in the liver, a process that is predomi-
nantly stimulated by glucagon. They are also used to synthesize
diacylglycerol which may contribute to hyperlipidemia and
increased very-low-density proteins (VLDL) [59]. Hyper-
glucagonemia results in ketogenesis via increased hepatic carni-
tine concentrations and decreased hepatic malonyl CoA, which
stimulates carnitine acyltransferase (CAT1), the rate-limiting
enzyme in ketogenesis. The clearance of ketone bodies is also
impaired in DKA due to low insulin concentrations, increased
glucocorticoids, and decreased peripheral glucose utilization [60].
510 M ET AB O LI S M CL IN I CA L A N D E XP E RI ME N TAL 6 5 ( 2 01 6 ) 5 0 7–52 1
Fig. 1 – Pathogenesis of DKA.
Growth hormone may also play a prominent role in
ketogenesis as physiological doses of growth hormone can
increase circulating levels of NEFA and ketone bodies. [61] The
ketoacids are buffered by extracellular and cellular buffers,
resulting in their loss and subsequent anion gap metabolic
acidosis. Elevated levels of other organic acids such as D-
lactate have been demonstrated in DKA [62,63]; D-lactate has
been shown to correlate with acidosis and anion gap in
subjects with DKA [62].
4.3. Water and Electrolyte Disturbances
Estimated electrolyte deficits in DKA are shown in Table 1.
Osmotic diuresis resulting from hyperglycemia promotes net
loss of multiple minerals and electrolytes such as sodium,
potassium, calcium, magnesium, chloride, and phosphate.
Some of these electrolytes (sodium, potassium and chloride)
can be replaced rapidly during treatment, while others may
require days or weeks to restore homeostasis [64–66]. Ketoanion
excretion results in obligate urinary cation excretion in the form
of sodium, potassium, and ammonium salts, which contributes
to a solute diuresis. Insulin deficiency per se may also contribute
to renal losses of water and electrolytes because of deficient
water and salt resorptive effect of insulin in the renal tubule [66].
Abnormalities in serum potassium are common in DKA
due to increased plasma tonicity, which results in intracellu-
lar water and potassium shifts into the extracellular space.
Also, protein catabolism with resultant potassium shifts into
the extracellular space, decreased potassium re-entry into the
cell secondary to insulinopenia and significant renal potassi-
um losses as a result of osmotic diuresis and ketonuria
contribute to potassium dyshomeostasis. Progressive volume
depletion leads to decreased glomerular filtration rate and a
greater retention of glucose and ketoanions in plasma which
elevates plasma tonicity. Thus, a considerable proportion of
patients with DKA have concomitant hypertonicity [37,67].
Patients with better oral intake of food, salt and fluid during
DKA have better preservation of kidney function, greater
ketonuria and lower ketonemia, lower anion gap and less
hypertonicity. During treatment with insulin, hydrogen ions
are consumed in the metabolism of ketoanion, this regener-
ates bicarbonate which relieves metabolic acidosis and
decreases the plasma anion gap. Therefore, the urinary loss
of ketoanions, as sodium and potassium salts, represents loss
of potential bicarbonate.
Emerging evidence indicate that hyperglycemic emergen-
cies including DKA are associated with an inflammatory state
marked by elevation in proinflammatory cytokines such as
 M E TAB O LI S M CL IN I CA L A N D EX PE RI ME N TA L 6 5 ( 2 01 6 ) 5 0 7–5 21 511

tumor necrosis factor-α, interleukins and C-reactive protein. ketones by dipstick, as well as initial arterial blood gases and
Also, reactive oxygen species, markers of lipid peroxidation complete blood count with differential. An electrocardiogram,
and plasminogen activator inhibitor-1 are elevated [68]. All of chest x-ray and urine, sputum or blood cultures may be
these parameters return to normal with resolution of DKA. clinically indicated. Glycated hemoglobin may be needed to
This inflammatory and procoagulant state may explain the differentiate an acute decompensation from a previously
relatively high incidence of thrombotic events in DKA. In-vivo undiagnosed or poorly controlled diabetes. The diagnostic
and in-vitro activation of T-cells with emergence of insulin criteria for DKA are shown in Table 1. Accumulation of
receptors has also been demonstrated [69,70]. ketoacids results in an increased anion gap metabolic
acidosis. The anion gap is calculated by subtracting the sum
of chloride and bicarbonate concentration from the sodium
5. Diagnosis concentration [Na − (Cl + HCO3)]. The normal anion gap was
previously reported to be 12 +/− 2 mEq/l, however, most
5.1. History and Physical Examination laboratories currently measure sodium and chloride concen-
trations using ion-specific electrodes, which measure plasma
DKA evolves rapidly over a short period, usually hours and chloride concentration 2–6 mEq/l higher than with prior
patients may not be aware of the disease. Symptoms of methods [77,78]. Thus, the normal anion gap using the
hyperglycemia such as polyuria, polydipsia, polyphagia and current methodology is between 7 and 9 mEq/l, therefore, an
weight loss are usually present. Other symptoms include anion gap >10–12 mEq/l may indicate the presence of anion gap
vomiting, abdominal pain, dehydration, weakness and in acidosis. DKA is classified as mild, moderate, or severe based on
severe cases altered mental status. Signs elicited on the the severity of metabolic acidosis (blood pH, bicarbonate,
physical examination include dehydration shown by poor ketones) and the presence of altered mental status [79].
skin turgor, Kussmaul respirations, and tachycardia. In Arterial blood gas analysis is recommended for the initial
severely ill patients, hypotension, shock and altered con- evaluation of patient with DKA, but this requires an expensive
sciousness may be present. Mental status can vary from full equipment which may not be available in some institutions;
alertness to profound lethargy or coma. The pathogenesis of especially in developing countries where the morbidity and
altered sensorium in DKA is not known with certainty, while mortality from DKA are high. Secondly, arterial blood sam-
some studies suggest hyperosmolarity as the origin of altered pling could be painful and technically difficult. Therefore,
mentation in DKA[67,71], others indicate acidosis may be the studies have investigated the use of venous blood to assess
prime determinant of the level of consciousness in these metabolic acidosis in subjects with DKA [79–82]. In a recent
patients [72,73]. However, in a retrospective analysis of over analysis of nearly 400 DKA cases we demonstrated that
200 cases of DKA we determined that acidosis was the prime arterial pH could be reliably estimated from serum bicarbon-
determinant of altered sensorium, but hyperosmolarity ate concentration using the following formula: arterial pH =
played a synergistic role in patients with severe acidosis to 6.97 + (0.0163 x bicarbonate), by applying this equation, serum
precipitate depressed sensorium. Combination of severe venous bicarbonate concentration of ≤ 20.6 mEq/L predicted
acidosis and hyperosmolarity predicted altered conscious-
ness with 61% sensitivity and 87% specificity [74]. Signs and
symptoms of the precipitating illness may be present and Table 2 – Admission biochemical data in patients with
should be sought in each case. Although infection is a DKA.
common precipitating factor fever may not be present, Parameters measured
patients can be normothermic or even hypothermic due to
Glucose (mg/dl) 616 ± 36
peripheral vasoconstriction arising from hypovolemia, and
Na + (mEq/l) 134 ± 1.0
low fuel substrate availability. Abdominal pain, which usually K+ (mEq/l) 4.5 ± 0.13
correlates with the severity of acidosis and may be confused BUN (mg/dl) 32 ± 3
with acute abdomen in 50–75% of cases [75,76]. In the absence Creatinine (mg/dl) 1.1 ± 0.1
of acidosis, another etiology for abdominal pain should be pH 7.12 ± 0.04
pursued. Hematemesis could occur in up to 25% of patients, Bicarbonate (mEq/l) 9.4 ± 1.4
ß-hydroxybutyrate (mmol/l) 9.1 ± 0.85
due to gastritis [57].
Total osmolality (mosm/kg) 323 ± 2.5
IRI (nmol/l) 0.07 ± 0.01
5.2. Laboratory Evaluation C-peptide (nmol/l) 0.21 ± 0.03
FFA (nmol/l) 1.6 ± 0.16
Table 2 summarizes the admission biochemical data in Human growth hormone (ng/l) 6.1 ± 1.2
patients with DKA. Cortisol (ng/l) 500 ± 61
A good clinical history and physical examination along IRI (nmol/l) ⁎ 0.09 ± 0.01
C-peptide (nmol/l) + 0.25 ± 0.05
with bedside tests such as capillary blood glucose and ketones
Glucagon (pg/l) 580 ± 147
or ketonuria should clinch a tentative diagnosis of DKA;
Catecholamines (ng/l) 1.78 ± 0.4
however, a definitive diagnosis must be verified by laboratory Anion gap 17
tests. The initial laboratory evaluation of patients with
Data are presented as mean ± SEM.
suspected DKA includes determination of plasma glucose,
⁎ Immunoreactive insulin.
blood urea nitrogen/creatinine, serum ketones, electrolytes +
Response to intravenous tolbutamide. Data adapted from ref [86].
(with calculated anion gap), osmolality, urinalysis, urine
512 M ET AB O LI S M CL IN I CA L A N D E XP E RI ME N TAL 6 5 ( 2 01 6 ) 5 0 7–52 1
arterial pH ≤ 7.3 with over 95% sensitivity and 92% accuracy.
Thus a gas analyzer may not be a necessity for managing
most patient with DKA [83]. Another cross-sectional prospec-
tive study evaluated the utility of end-tidal carbon dioxide
measured by capnography as a surrogate for arterial blood gas
in 181 patients suspected to have DKA in the emergency
room. Capnography values higher than 24.5 mmHg excluded
DKA with a sensitivity and specificity of 90% [84].
Patients with DKA could have leukocytosis, which may be
proportional to the severity of acidosis, hypercortisolemia
and elevation of catecholamines [85]. However, leukocytosis
greater than 25,000/μL may suggest a concurrent infection
which warrants further evaluation [86]. The admission serum
sodium is usually low because of the efflux of water from the
intracellular to the extracellular space in the presence of
hyperglycemia. An increase in serum sodium concentration
in the presence of hyperglycemia indicates a rather profound
degree of water loss. The measured value can be corrected by
adding 1.6 mmol/l (1.6 mEq/L) of sodium for every 5.6 mmol/l
(100 mg/dl) of glucose above 5.6 mmol/l (100 mg/dl). Measured
serum sodium and glucose concentrations may be falsely
lowered by severe hypertriglyceridemia in laboratories using
volumetric testing or dilution of samples with ion–specific
electrodes [87,88]. Unless the plasma is cleared of chylomicrons,
pseudonormoglycemia and pseudohyponatremia can occur.
Serum potassium concentration may be elevated because
of an extracellular shift of potassium caused by insulin
deficiency, hypertonicity, and acidemia [89,90]. Patients with
low normal or low serum potassium concentration on
admission have severe total-body potassium deficiency and
require very careful cardiac monitoring and more vigorous
potassium replacement as treatment can lower serum potas-
sium further, thus predisposing to cardiac dysrhythmia. The
total body deficit of sodium and potassium might be as high
as 500–700 mEq [90,91]. Therefore, treatment with insulin
should not be commenced until the serum potassium is
greater than or equal to 3.3 mmol/l (3.3 mEq/L). Measured
serum creatinine may be falsely elevated due to dehydration
or interference with the assay by elevated acetoacetate
[92,93]. Serum creatinine should be monitored during hydra-
tion and resolution of acidosis.
A significant proportion of patients with DKA may have
elevated plasma tonicity, which in the presence of severe
acidosis contributes to altered mentation; about 7% of
patients may have severe acidosis and hypertonicity and are
therefore at the risk of developing altered state of conscious-
ness and poorer prognosis [74]. Perhaps the term Diabetic
hyperosmolar ketoacidosis (DHKA) could be applied to this
subset of patients. In the calculation of effective osmolality
[2[measured Na (mEq/L)] + glucose (mg/dl)/18], the urea
concentration is not taken into account because it is freely
permeable and its accumulation does not induce major
changes in intracellular volume or osmotic gradient across
the cell membrane. Elevation in amylase and lipase levels
may occur in 16–25% of patients with DKA [94,95]. Although
amylase level may correlate with the severity of acidosis in
such patients [95], its origin may be the parotid gland [94].
Therefore, pancreatic enzymes may not be specific for the
diagnosis of pancreatitis in patients with DKA [96]; hence,
patients with elevated serum amylase and lipase and
abdominal pain may require imaging to exclude pancreatitis.
Ketone bodies are usually measured by the nitroprusside method
which detects acetoacetate but not ß hydroxybutyrate (BOHB),
the more abundant ketone body. During treatment of DKA, BOHB
is converted to acetoacetate, therefore, the follow-up measure-
ment of ketones with the use of nitroprusside is not recom-
mended because the ketones test may show high values which
erroneously suggests that the condition of ketonemia is worsen-
ing. Newer glucose meters have the capability to measure BOHB,
which overcomes this problem. Furthermore, studies have
investigated the use of BOHB in the management of DKA and
this metabolite is increasingly being used by laboratories to
evaluate ketonemia [96].
Drugs that have sulfhydryl groups can interact with the
reagent in the nitroprusside reaction, thus giving a false
positive result [97], captopril, an angiotensin converting
enzyme inhibitor used for the treatment of hypertension
and diabetic nephropathy is an important example. Clinical
judgment and other biochemical considerations may be
relevant in interpreting the value of positive nitroprusside
reaction in patients taking captopril who are suspected to
have DKA.
5.3. Differential Diagnosis
DKA consists of the biochemical triad of hyperglycemia,
ketonemia and anion gap metabolic acidosis, each of these
components can be seen in other metabolic conditions (Fig 2).
Therefore, physical and laboratory evaluation to differentiate
other causes of metabolic acidosis is warranted. For example,
in alcoholic ketoacidosis (AKA), total ketone bodies are much
greater than in DKA with a higher BOHB to acetoacetate ratio
of 7:1 versus a ratio of 3:1 in DKA [98]. Patients with AKA
seldomly present with hyperglycemia. Patients with poor oral
intake could present with mild ketoacidosis, (starvation
ketosis), but this may not occur in individuals on prolonged
fasting, except they have a problem with ketone metabolism.
Thus, patients with starvation ketosis rarely present with
serum bicarbonate concentration less than 18 mEq/L, and do
not exhibit hyperglycemia. DKA should be distinguished from
high anion gap acidosis including lactic acidosis, advanced
chronic renal failure as well as ingestion of drugs such as
salicylate, methanol and ethylene glycol. Isopropyl alcohol
which is commonly available as rubbing alcohol can cause
considerable ketosis and high osmolar gap without metabolic
acidosis, however, it has a tendency to cause hypoglycemia
rather than hyperglycemia. These conditions with their
laboratory findings are summarized in Table 3 [99].
6. Treatment
The goals of therapy in DKA are 1) Improvement of circulatory
volume and tissue perfusion; 2) Gradual correction of hyperglycemia
and hyperosmolality; 3) Correction of electrolyte imbalance, and
resolution of ketosis; 4) Identification and adequate treatment of co-
morbid conditions. The recommended protocol for the treatment of
DKA is provided in Fig. 3 [78,100]. Successful treatment of DKA
demands frequent monitoring by clinical and laboratory parameters
to ensure the goals of therapy are being achieved.
 M E TAB O LI S M CL IN I CA L A N D EX PE RI ME N TA L 6 5 ( 2 01 6 ) 5 0 7–5 21
Hyper-
glycemia
Other Hyperglycemic
States
Uncontrolled DM DKA
HHS
Stress hyperglycemia
Ketosis
Other Ketotic States.
Ketotic hypoglycemia
Alcoholic ketosis
Starvation ketosis
Isopropyl alcohol
Hyperemesis
Fig. 2 – Differential diagnosis of DKA. Data adapted from ref [19].
6.1. Fluids
DKA is a volume-depleted state [89,101] water deficit may be
up to 6 l (Table 1); Initial fluid therapy is aimed at expanding
interstitial and intravascular volume and reestablishing
adequate renal perfusion. Although the benefits of proper
rehydration is unequivocal, the fluid of choice for resuscitat-
ing the critically ill patient was a subject of controversy. A
prospective study which investigated the effects of hypotonic,
isotonic and hypertonic fluids in patients with severe diabetic
ketoacidosis reported no significant difference in the volume
of fluid retained when solutions of varying tonicity were
administered. However, hypertonic fluids worsened hyperto-
nicity, hypernatremia and hyperchloremia [102]. Additionally,
some patients treated with hypotonic fluids developed
diuresis; hence, rapid repletion of the plasma and extracellu-
lar volume with isotonic fluids is indicated in subjects with
DKA. Furthermore, there was controversy as to which fluid is
superior in the critically ill: colloids such as dextran vs
crystalloids such as normal saline. A meta-analysis of
prospective randomized studies comparing colloids and
crystalloids in critically ill patients reported that colloids
were more expensive but did not confer any mortality benefit
[103]. Another prospective randomized study investigated the
optimal rate of hydration in patients with DKA. Subjects were
randomized to receive either 1000 ml/h or 500 ml/h of 0.9%
saline solution. Both groups were biochemically similar at
baseline and in the rate of resolution of biochemical abnor-
malities [104]. Lastly, a prospective randomized study which
evaluated fluids for maintenance of adequate glycemic level
for the resolution of DKA (5% versus 10% dextrose along with
continued insulin infusion), found that 10% dextrose was
associated with significantly lower level of ketonemia and
higher level of hyperglycemia, but did not confer any
advantage in the rate of resolution of acidosis [105].
513
Acidosis
Other Metabolic Acidotic
States
Lactic acidosis
Hyperchloremic acidosis
Salicylism
Adapted from ref 19
The initial fluid of choice is isotonic saline which should be
infused at the rate of 15–20 ml/kg body weight per hour or 1–
1.5 l during the first hour. This expands the extracellular
volumes. The subsequent choice for fluid replacement depends
on the state of hydration, serum electrolyte levels, and urinary
output. In general, 0.45% NaCl infused at 4–14 ml/kg/h is
appropriate if the corrected serum sodium is normal or
elevated; 0.9% NaCl at a similar rate is appropriate if corrected
serum sodium is low (See Fig. 3). Adequate fluid replacement is
assessed by hemodynamic monitoring (improvement in
blood pressure and pulse), measurement of fluid input and
output, blood chemistry and clinical examination. Half of the
estimated water deficit should be replaced over 12–24 h. An
effective serum osmolality >320 mOsm/kg indicates severe
dehydration, which would require aggressive fluid replacement
therapy. In patients with hypotension, aggressive fluid reple-
tion with isotonic saline should continue until blood pressure is
stable. The administration of insulin without fluid replacement
in hypotensive patients could worsen circulatory collapse.
Hydration in the first hour of therapy before insulin has the
following advantages: a) it allows opportunity to obtain serum
potassium value before insulin administration, b) it prevents
potential deterioration of hypotension, which could occur with
the use of insulin without adequate hydration, c) it improves
insulin action [106] and may reduce the concentration of
counter regulatory hormones and hyperglycemia [107]. Hydra-
tion has been shown to reduce blood glucose, BUN,
hemoconcentration and potassium concentration without
significant reduction in pH or HCO3 concentration [67]. Reduc-
tion in blood glucose level is thought to result from osmotic
diuresis [93,107]. Energy is required for the metabolism of
ketone bodies, therefore, as soon as blood glucose falls below
200 mg/dl, the sodium chloride solution should be replaced
with 5% glucose containing saline solution and the rate of
insulin infusion should be reduced until acidosis and ketosis
514 M ET AB O LI S M CL IN I CA L A N D E XP E RI ME N TAL 6 5 ( 2 01 6 ) 5 0 7–52 1

Isoproproply resolve. It should be emphasized that the replacement

Negative
of urinary losses is also important as failure to do this leads

Normal

Normal
alcohol

↑↑
to delay in the restoration of sodium, potassium, and

↑
water deficits.
Rhabdomyolysis

6.2. Insulin Therapy

Mild ↓ may be ↓↓

Myoglobinuria
or slight ↑
Negative
Normal

Normal

Normal
DKA was invariably fatal before the discovery of insulin in the

↑↑

↑
early 1920s; with the introduction of insulin, mortality was
reduced to less than 50% and subsequently to less than 20% as
antibiotics and adequate hydration were incorporated into
Hypoglycemic

the treatment of DKA [3,108]. In the 1950s, the mortality of

hemoglobinuria
<30 mg/dl

or slight ↑

patients treated with high doses of insulin was reported to be

Negative

or slight
Normal

Normal

Normal

Normal

Normal
less than 10% but in more recent years, the use of standard-
↓↓
coma

ized treatment guidelines has reduced mortality rate to less

than 1% in good centers [3,108]. In the beginning of insulin
therapy, small amounts of insulin were used to treat DKA
>330 mOsm/kg
>500 mg/dl
Hyperosm-

with good results, but high-dose insulin therapy became the

or slight ↑
olar coma

Normal

Normal

Normal

Normal

standard of care when insulin became more available;

++
↑↑

↑↑

between the 1950s and early 1970s, up to 100 U/h or more

were given due to perceived insulin resistance, but prospec-
tive randomized studies did not show any advantage of high-
ethylene glycol

Serum positive

dose insulin compared with low-dose [108]. Landmark studies

Methanol or

intoxication

Negative
Normal

Normal

Normal

in the 1970s established low or physiologic dose regular

↑↑
↓

insulin as the optimal therapy for DKA [109,110]. A prospec-

tive randomized controlled trial by Kitabchi et al. investigated
the effect of low-dose vs high-dose insulin therapy in 48
intoxication

patients with DKA [110,110]. The baseline biochemical char-

Serum levels
Normal or ↓

Negative †

Normal or
Salicylate

positive
Normal

Normal

acteristics were similar in the two arms, which were

↓↑

comparable in the rate of resolution of biochemical defects

and the counter-regulatory hormones glucagon and cortisol
declined at the same rate in both groups. However, hypogly-
cemia and hypokalemia were more frequent in subjects
↓ or normal

moderate ↑

salicylate
(starvation)

Negative
Slight to

Normal

treated with high-dose vs physiologic dose insulin (25 vs 0%

Alcoholic

↓↑

and 30 vs 4% respectively).
ketosis

In moderate to severe DKA or DKA with mental

obtundation (as defined in Table 1), intravenous regular
Table 3 – Laboratory evaluation of metabolic acidosis and coma.

insulin by continuous infusion is the treatment of choice.

Negative
acidosis

Normal

Normal

Normal
Slight ↑
Uremic

Mild ↓

mg/dl
BUN

Previous treatment protocols have recommended the admin-

↑

istration of an initial bolus of 0.1 U/kg followed by the infusion

of 0.1 U/kg/h [19,85], however, a more recent prospective
May give lactate Serum >200
for ethylene glycol >7 mmol/l

randomized trial demonstrated that a bolus is not necessary if

Negative
Normal

Normal

Normal

Normal
acidosis

patients are given hourly insulin infusion at 0.14 U/kg body

Lactic

wt/h [100]. Low-dose insulin infusion protocols decrease

plasma glucose concentration at a rate of 50–75 mg/dL/h. If
blood glucose does not fall by 10% in the first hour, an
intravenous bolus of 0.14 U/kg should be administered
followed by continuous infusion at the previous rate [78].
++

↑↑
↓

When the plasma glucose reaches 200 mg/dl, the insulin

DKA

infusion rate should be reduced to 0.02–0.05 U/kg/h. Also,

dextrose should be added to the intravenous fluids at this point.
Starvation

(starvation)
fat intake

The rate of insulin administration or the concentration of

Negative

positive
Normal

Normal

Normal
Slight ↑

Slight ↑

False-
or high

Mild

dextrose may be adjusted to maintain blood glucose concentra-

tion between 150 and 200 mg/dl until resolution of DKA.
Several prospective randomized open label trials have
Total plasma

Miscellaneous

demonstrated the efficacy and cost effectiveness of subcuta-

Osmolality
Glycosuria

Anion gap

neous rapid-acting insulin analogs (lispro, aspart and

Uric Acid
ketones*
glucose
Plasma

glulisine) in the treatment of uncomplicated mild to moderate

DKA [111–115]. In one of these studies [111], the patients
pH

received subcutaneous insulin lispro at a dose of 0.3 U/kg

 M E TAB O LI S M CL IN I CA L A N D EX PE RI ME N TA L 6 5 ( 2 01 6 ) 5 0 7–5 21
Fig. 3 – Treatment of DKA.
initially, followed by 0.1 U/kg every hour until blood glucose
was < 250 mg/dl, when insulin dose was decreased to 0.05 or
0.1 U/kg given every hour until resolution of DKA. In the
second study [112], insulin aspart was given at an initial dose
of 0.3 U/kg followed by 0.1 U/kg every hour until blood glucose
was <250 mg/dl when dose was reduced to 0.05 U/kg hourly
until resolution of DKA. The second arm of the aspart study
administered 0.3 U/kg as a loading dose followed by 0.2 U/kg
an hour later and every 2 h until blood glucose was <250 mg/dl.
At that time, insulin dose was reduced to 0.1 U/kg every 2 h.
In comparison with intravenous regular insulin, there were
no differences in the length of hospital stay, total amount of
insulin needed for resolution of hyperglycemia or
ketoacidosis. Patients treated with insulin analogs were
managed in the open medical wards which reduced cost of
hospitalization by 30%. In another study which investigated
the efficacy of insulin analogs, patients were randomly
assigned to receive intravenous regular or glulisine insulin
until resolution of DKA [114]. Thereafter, patients treated
with regular insulin were transitioned to subcutaneous NPH
and regular insulin given twice daily, while subjects treated
with glulisine insulin were transitioned to subcutaneous
glargine given once daily and prandial glulisine. Both arms
were equally effective but patients treated with NPH and
regular insulin had a higher rate of hypoglycemia than the
glargine and glulisine group (41% vs 15%). Insulin analogs
have not been investigated in complicated or severe DKA,
therefore, it would be prudent to treat patients with severe
515
DKA, hypotension, anasarca, or associated severe critical
illness with intravenous regular insulin in the ICU [78].
Patients with DKA should be treated with regular insulin or
insulin analog until resolution. Criteria for resolution of
ketoacidosis include a blood glucose < 200 mg/dl and two of
the following criteria: a serum bicarbonate level ≥ 15 mEq/l, a
venous pH > 7.3, and a calculated anion gap ≤ 12 mEq/l.
Patients can be transitioned to subcutaneously administered
multiple dose insulin when DKA has resolved and they are
able to eat. Those previously treated with insulin may be
recommenced on their home dose if they had been well
controlled. Insulin-naïve patients should receive a multi-dose
insulin regimen beginning at the dose of 0.5–0.8 U/kg/day [78].
To prevent recurrence of ketoacidosis in the transition period,
insulin infusion should be discontinued 2 h after commence-
ment of subcutaneous insulin. If the patient is unable to eat,
intravenous insulin and fluid replacement may be continued.
Use of long-acting insulin analogs during the initial manage-
ment of DKA may facilitate transition from intravenous to
subcutaneous insulin therapy. It may also avoid rebound
hyperglycemia and ketogenesis, which could occur on dis-
continuation of intravenous insulin [116]. This hypothesis,
which looks plausible has not been verified by prospective
studies. A recent pilot study investigated this approach in a
prospective, randomized, controlled trial comparing coad-
ministration of insulin glargine and intravenous insulin vs
intravenous insulin alone [117]. The outcome was similar in
both arms of the study. All patients received intravenous
516 M ET AB O LI S M CL IN I CA L A N D E XP E RI ME N TAL 6 5 ( 2 01 6 ) 5 0 7–52 1
insulin while the glargine arm received subcutaneous insulin
glargine in addition within two hours of diagnosis. Upon
resolution of DKA, patients treated with intravenous insulin
alone were transitioned to long-acting insulin while daily
insulin glargine was continued in the glargine group. During
the follow up period, blood should be drawn every 2–4 h for
determination of serum electrolytes, glucose, blood urea
nitrogen, creatinine, osmolality, and pH. Determination of
pH can be done by venous rather than arterial puncture
because arterial blood gases are seldomly needed in such
patients. An equivalent arterial pH value is calculated by
adding 0.03 to the venous pH value [80]. In facilities where
blood gas analysis is not available, venous serum bicarbonate
could be used to estimate arterial pH. [83]. Ketonemia typically
takes longer to resolve than hyperglycemia. Direct measurement
of BOHB in the blood is the preferred method for monitoring DKA
and is becoming increasingly available [118].
6.3. Potassium replacement
Close attention should be paid to repletion of potassium in
patients with DKA. Although total-body potassium is depleted
[119,120], mild to moderate hyperkalemia is frequently seen
in these patients, due to acidosis which displaces potassium
from the cell to the extracellular space, proteolysis and
insulin deficiency [85]. Insulin therapy, correction of acidosis,
and volume expansion would decrease serum potassium
concentration. To prevent hypokalemia, potassium replace-
ment should be initiated when serum levels fall below
5.3 mEq/l in patients with adequate urine output (50 ml/h).
Generally, 20–30 mEq potassium in each liter of infusion fluid
is sufficient to maintain a serum potassium concentration
within the normal range of 4–5 mEq/L. Occasionally patients
with DKA may present with hypokalemia, especially if they
have been vomiting or had been taking diuretics. In such
cases, potassium repletion should begin with fluid therapy,
and insulin treatment should be delayed until potassium
concentration is restored to > 3.3 mEq/L to avoid arrhythmias
or cardiac arrest and respiratory muscle weakness [78]. A
prospective study of 29 consecutive cases of DKA reported
normokalemia or hyperkalemia in 82% of the patients, but
63% of them developed hypokalemia in course of treatment,
the correction of which required 145 mEq of potassium on the
average (59–239 mEq) [102]. Cardiac monitoring would be
indicated in subjects with severe hypokalemia.
6.4. Bicarbonate therapy
Bicarbonate therapy in DKA is controversial. While some workers
believe that insulin therapy would correct ketoacidosis without
use of bicarbonate, others argue that severe acidosis is associated
with impaired myocardial contractility, cerebral vasodilatation,
coma, and gastrointestinal sequelae, which warrants bicarbonate
therapy. Prospective randomized studies have demonstrated
no benefits of bicarbonate therapy in DKA patients with pH ≥6.9
[121–123]. In a randomized study of 21 adults [121], administration
of bicarbonate conferred no benefits in the rate of resolution of
DKA or increase in pH or serum bicarbonate concentration in the
blood or cerebrospinal fluid. Furthermore, it was observed that the
pH level was higher in the CSF than the blood indicating the brain
has some protection against severe acidosis. In another random-
ized study of 32 patients, bicarbonate therapy was associated with
delay in the fall of total ketone bodies, blood lactate and lactate:
pyruvate ratio [123]. Delay in the resolution of ketosis was also
observed in human and animal experiments in another small
prospective study [124]. Lastly, a more recent systematic review
which investigated the benefits and risks of bicarbonate therapy
in DKA reported increased risk of cerebral edema and prolonged
hospitalization in pediatric patients but no benefit [125]. No
prospective randomized studies investigating the use of bicar-
bonate in DKA with pH values <6.9 have been reported [78],
therefore, the decision to use bicarbonate should be based on the
clinical state of the patient. Subjects who are clinically well
compensated may not require administration of bicarbonate,
while it would be prudent to use bicarbonate in individuals with
severe acidosis who may deteriorate without bicarbonate therapy.
Adults with pH <6.9 who may deteriorate without bicarbonate
therapy may be given 100 mmol sodium bicarbonate in 400 ml
sterile water (an isotonic solution) with 20 mEq KCI administered
at a rate of 200 ml/h for 2 h until the venous pH is >7.0. If the pH is
still <7.0 after infusion, we recommend repeating infusion every
2 h until pH reaches >7.0 [78].
6.5. Phosphate therapy
Phosphate moves along with potassium from the intracellular to
the extracellular compartment in response to acidosis and
osmotic diuresis leads to urinary phosphate loss. Although,
whole body phosphate deficits may average about 1.0 mmol/kg
of body weight in DKA, serum phosphate at presentation is
typically normal or high due to shifts across the cell membrane
[126–128]. During insulin therapy, phosphate re-enters the
intracellular compartment leading to a fall in serum phosphate
concentrations. Adverse complications of hypophosphatemia
are uncommon but could occur in severe cases. Potential
complications include respiratory and skeletal muscle weakness,
hemolytic anemia, and poor cardiac performance [128]. Phos-
phate depletion may also contribute to decreased concentrations
of 2,3-diphosphoglycerate (2,3-DPG), thus shifting the oxygen
dissociation curve to the left and limiting tissue oxygen delivery
[128]. Prospective randomized studies have shown no benefit
of phosphate replacement on clinical outcome in DKA [126,127]
and overzealous phosphate therapy may be associated
with hypocalcaemia [127]. To avoid the sequalae of
hypophosphatemia, careful phosphate replacement may be
indicated in patients with cardiac dysfunction, anemia, respira-
tory depression, and in those with serum phosphate concentra-
tion lower than 0.32 mmol/l (1.0 mg/dl). To replace phosphate
stores and to avoid hyperchloremia, intravenous potassium
repletion can be administered in a ratio of two-thirds potassium
chloride and one-third potassium phosphate. The maximal rate
of phosphate replacement generally regarded as safe to treat
severe hypophosphatemia is 4.5 mmol/h (1.5 ml/h of K2PO4) [129].
7. Complications
The most frequent complications of DKA are hypoglycemia and
hypokalemia, which result from overzealous treatment with
insulin. Hypokalemia may also complicate bicarbonate therapy.
 M E TAB O LI S M CL IN I CA L A N D EX PE RI ME N TA L 6 5 ( 2 01 6 ) 5 0 7–5 21
A transient hyperchloremic non-anion gap acidosis could occur
in the recovery phase of DKA [130,131]; due to the loss of large
quantities of ketoanions. Ketoanions are metabolized with
regeneration of bicarbonate, their loss in the urine hinders
regeneration of bicarbonate during treatment thus predisposing
to acidosis. Administration of intravenous fluids containing
chloride that exceeds the plasma chloride concentration and
the intracellular shifts of NaHCO3 during correction of DKA also
contribute to hyperchloremic acidosis [131]. This complication is
usually of little clinical consequence.
Cerebral edema is a rare but serious complication of DKA,
occurring in 0.7–1.0% of children with DKA; especially in those with
newly diagnosed diabetes. Cerebral edema can also occur in
patients with previously diagnosed diabetes and in very young
adults under 20 years of age [132,133]. Headache, the earliest
symptom of cerebral edema is followed by lethargy and altered
sensorium. Neurological deterioration may occur rapidly with
seizures, incontinence, pupillary changes, bradycardia, and respi-
ratory arrest setting in as brain stem herniation occurs. Papilledema
may be absent if onset is rapid. Mortality rate may be >70% if
neurological symptoms are established, with only 7–14% of patients
recovering without sequalae. Postulated mechanisms for cerebral
edema include osmotically-driven movement of water into the
central nervous system when plasma osmolality declines too
rapidly during the treatment. [132–135]. A study, which assessed
cerebral water diffusion and cerebral vascular perfusion during the
treatment of children with DKA using magnetic resonance imaging,
reported that cerebral edema was due to increased cerebral
perfusion [134]. Another putative mechanism for cerebral edema
involves the cell membrane Na+/H+ exchanger, which is activated
in DKA. The high H+ level allows more Na + into the cell, which
attracts water into the cell leading to edema [136]. The ketone
bodies acetoacetate and β-hydoxybutyrate may also play a role in
the pathogenesis of cerebral edema [137]. Ketone bodies have been
shown to affect vascular integrity and permeability and contribute
to edema formation. Measures that may reduce the risk of cerebral
edema in high-risk patients include gradual replacement of sodium
and water deficits in patients who have hypertonicity and the
addition of dextrose to the correction fluid once blood glucose
reaches 250 mg/dl.
Hypoxemia and, rarely, noncardiogenic pulmonary edema
may complicate the treatment of DKA. Hypoxemia is due to a
reduction in colloid osmotic pressure that results in increased
lung water content and decreased lung compliance [19]. Patients
with DKA who have a widened alveolo-arteriolar oxygen gradient
noted on initial blood gas measurement or with pulmonary rales
on physical examination appear to be at higher risk for the
development of pulmonary edema. Thrombosis including dis-
seminated intravascular coagulation has been reported in DKA
[138] and high levels of pro-inflammatory cytokines and plasmin-
ogen activator inhibitor-1 (PAI-1) have been demonstrated in DKA,
which resolve with insulin therapy and correction of hyperglyce-
mia [68], therefore, prophylactic use of heparin may be indicated.
8. Prevention
Although mortality from DKA is < 1%, all course mortality in
subjects surviving DKA may be high; a retrospective analysis
of over 600 episodes of DKA in nearly 300 patients seen in a
517
hospital in the UK from 2007 to 2012 recorded no inpatient
mortality. However, 14% of the patients died within 5 years of
follow up and mortality was higher in subjects who had
recurrent DKA [139]. Predictors of mortality during follow up
included psychological issues, peripheral neuropathy, ische-
mic heart disease, alcoholism and prior admission to inten-
sive care. Therefore, measures to prevent recurrent DKA
should be pursued vigorously. Subjects with diabetes have
increased cardiovascular morbidity and mortality; although
SGLT-2 inhibitors have been associated with DKA,
empagliflozin, an SGLT-2 inhibitor was recently reported to
reduce all-cause and cardiovascular mortality in patients with
type 2 diabetes at high risk for cardiovascular events [140].
Prospective clinical studies have identified poor adherence to
insulin therapy as the major precipitant of DKA in some
populations [141,142]. Education of the patient and care giver
about diabetes care especially sick day management would be
beneficial in preventing DKA. An interventional study in teen
age patients with type 1 diabetes, which incorporated
frequent outpatient clinics, reported better diabetes control
and less diabetes related hospitalization in the intervention
group [143]. Again, a home-based psychotherapy program
reduced hospital admission for DKA over 24 months in a
prospective randomized study of 127 youths [144]. Illicit drug
use is associated with recurrent DKA [24,145,146], therefore,
rehabilitation should be helpful in drug addicts. A significant
proportion of DKA occur in subjects with type 2 diabetes; adoption
of healthy lifestyle and maintenance of ideal body weight would
contribute in reducing the surging incidence of DKA.
9. Future perspective
Over the years, remarkable improvement has been made in the
prognosis of DKA, however, the recent observation of increased
all course mortality in survivors of DKA merits a prospective
investigation. Furthermore, some questions remain to be an-
swered; the use of bicarbonate in patients with pH < 6.9 is yet to
be investigated in a prospective randomized study. The mecha-
nism for the induction of proinflammatory and prothrombotic
state in DKA remains unclear. Understanding this pathway may
be useful in preventing cardiovascular morbidity associated with
hyperglycemic crises. Fast-acting insulin analogs have been
shown to be as effective as intravenously administered regular
insulin in mild to moderate DKA, but it is not known if
subcutaneously administered regular insulin would be equally
efficacious in such patients. Using regular insulin by subcutane-
ous route would be more economical than insulin analogs
especially in developing countries. The rising prevalence of DKA
could be due to its incidence in patients with ketosis-prone type 2
diabetes. The mechanism for acute severe β-cell function failure
leading to ketoacidosis remains a subject for further investiga-
tion. The pathogenesis of altered mentation in DKA is yet to be
conclusively elucidated. A retrospective study suggests acidosis
may be the major determinant of sensorium [74], but a
prospective randomized study is needed to validate this obser-
vation. The role of other organic acids such as lactate in the
pathogenesis of DKA may also need to be evaluated further.
Lastly, recent reports of DKA in subjects treated with SGLT2
inhibitors certainly deserves further inquiry.
518 M ET AB O LI S M CL IN I CA L A N D E XP E RI ME N TAL 6 5 ( 2 01 6 ) 5 0 7–52 1
Contribution of Authors
EAN – draft of manuscript and review for intellectual content;
AEK – critique of manuscript and review for intellectual content.
Funding and Conflicts of interest
None.
Acknowledgment
Administrative assistance by Ms. Tara Bea is greatly appreciated.
REFERENCES
[1] Tattersall RB. The history of diabetes mellitus. In: RIG H, CS
C, A F, BJ G, editors. Textbook of diabetes. 4th ed. West
Sussex, UK: Wiley-Blackwell; 2010. p. 3–23.
[2] Dreschfeld J. The Bradshawe Lecture on Diabetic Coma. Br
Med J 1886;2(1338):358–63.
[3] Dave J, Chatterjee S, Davies M, et al. Evaluation of
admissions and management of diabetic ketoacidosis in a
large teaching hospital. Pract Diab Int 2004;21:149–53.
[4] Wang ZH, Kihl-Selstam E, Eriksson JW. Ketoacidosis occurs
in both Type 1 and Type 2 diabetes- a population-based
study from Northern Sweden. Diabet Med 2008;25:867–70.
[5] Faich GA, Fishbein HA, Ellis SE. The epidemiology of diabetic
acidosis: a population-based study. Am J Epidemiol 1983;
117:551–8.
[6] Centers for Disease Control and Prevention. National Hospital
Discharge Survey (NHDS). Available at: www.cdc.gov/nchs/
about/major/hdasd/nhds.htm, [accessed 1/20/2009)].
[7] Graves EJ, Gillium BS. The National Center for Health Statistics.
Detailed diagnoses and procedures: National Hospital
Discharge Survey. Vital Health Stat 13 1997;130:1–146.
[8] Malone ML, Gennis V, Goodwin JS. Characteristics of
diabetic ketoacidosis in older versus younger adults. J Am
Geriatr Soc 1992;40:1100–4.
[9] Wolfsdorf J, Glaser N, Sperling MA. Diabetic ketoacidosis in
infants, children, and adolescents: a consensus statement
from the American Diabetes Association. Diabetes Care
2006;29:1150–2259.
[10] White NH. Diabetic ketoacidosis in children. Endocrinol
Metab Clin North Am 2000;29:657–82.
[11] Mbugua PK, Otieno CF, Kayima JK, et al. Diabetic
ketoacidosis: clinical presentation and precipitating factors
at Kenyatta National Hospital. East Afr Med J 2005;82:S191–6.
[12] Elmehdawi RR, Ehmida M, Elmahrehi H, et al. Incidence and
mortality of diabetic ketoacidosis in Benghazi-Libya in 2007.
Med J 2013;3:178–83.
[13] National Center for Health Statistics, CDC, Agency for
Healthcare Research and Quality. Databases and related
tools from the healthcare cost and utilization project
(HCUP); 2009[Available at www.hcup-us.ahrq.gov/reports/
statbriefs accessed January 20,].
[14] Munro JF, Campbell IW, McCuish AC, et al. Euglycaemic
diabetic ketoacidosis. Br Med J 1973;2(5866):578–80.
[15] Burge MR, Hardy KJ, Schade DS. Short-term fasting is a
mechanism for the development of euglycemic ketoacidosis
during periods of insulin deficiency. J Clin Endocrinol Metab
1993;76(5):1192–8.
[16] Peters AL, Buschur EO, Buse JB, et al. Euglycemic diabetic
ketoacidosis: A potential complication of treatment with
sodium–glucose cotransporter 2 inhibition. Diabetes Care
2015 [Epub ahead of print].
[17] Hine J, Paterson H, Abrol E, et al. SGLT inhibition and
euglycaemic diabetic ketoacidosis. Lancet Diabetes
Endocrinol 2015;3(7):503–4.
[18] Umpierrez GE, Smiley D, Kitabchi AE. Ketosis-prone type 2
diabetes mellitus. Ann Int Med 2006;144:350–7.
[19] Kitabchi AE, Nyenwe EA. Hyperglycemic crises in diabetes
mellitus: diabetic ketoacidosis and hyperglycemic
hyperosmolar state. Endocrinol Metab Clin North Am 2006;
35:725–51.
[20] Newcomer JW. Second generation (atypical) antipsychotics
and metabolic effects: a comprehensive literature review.
CNS Drugs 2005;19(Suppl. 1):1–93.
[21] Branis NM, Wittlin SD. Amphetamine-like analogues in
diabetes: Speeding towards ketogenesis. Case Rep
Endocrinol 2015;2015:917869. http://dx.doi.org/10.1155/2015/
917869 [Epub 2015 Apr 19].
[22] Taylor SI, Blau JE, Rother KI. Perspective: SGLT2 inhibitors
may predispose to ketoacidosis. J Clin Endocrinol Metab
2015;18:jc20151884 [Epub ahead of print].
[23] Polonsky WH, Anderson BJ, Lohrer PA, et al. Insulin omission in
women with IDDM. Diabetes Care 1994;17:1178–85.
[24] Nyenwe EA, Loganathan RS, Blum S, et al. Active use of
cocaine: an independent risk factor for recurrent diabetic
ketoacidosis in a city hospital. Endocr Pract 2007;13:22–9.
[25] Isidro ML, Jorge S. P recreational drug abuse in patients
hospitalized for diabetic ketosis or diabetic ketoacidosis.
Acta Diabetol 2013;50:183–7.
[26] Shah P, Isley WL. Ketoacidosis during a low carbohydrate
diet. N Engl J Med 2006;354:97–8.
[27] Peden NR, Broatan JT, McKenry JB. Diabetic ketoacidosis
during long-term treatment with continuous subcutaneous
insulin infusion. Diabetes Care 1984;7:1–5.
[28] Katz JR, Edwards R, Kahn M, et al. Acromegaly presenting
with diabetic ketoacidosis. Postgrad Med J 1996;72:682–3.
[29] Edelman ER, Stuenkel CA, Rutherford JD, et al. Diabetic
ketoacidosis associated with pheochromocytoma. Cleve
Clin J Med 1992;59:423–7.
[30] Isotani H, Fujimura Y, Furukawa K, et al. Diabetic
ketoacidosis associated with the pheochromocytoma of
youth. Diabetes Res Clin Pract 1996;34:57–60.
[31] Dodu SR. Diabetes in the tropics. Br Med J 1967;2:747–50.
[32] Adadevoh BK. “Temporary diabetes” in adult Nigerians.
Trans R Soc Trop Med Hyg 1968;62:528–30.
[33] Winter WE, Maclaren NK, Riley WJ, et al. Maturity-onset diabetes
of youth in black Americans. N Engl J Med 1987;316(6):285–91.
[34] Maldonado M, Hampe CS, Gaur LK, et al. Ketosis-prone
diabetes: dissection of a heterogeneous syndrome using an
immunogenetic and beta-cell functional classification, pro-
spective analysis, and clinical outcomes. J Clin Endocrinol
Metab 2003;88:5090–8.
[35] Mauvais-Jarvis F, Sobngwi E, Porcher R, et al. Ketosis-prone
type 2 diabetes in patients of sub-Saharan African origin:
clinical pathophysiology and natural history of beta-cell
dysfunction and insulin resistance. Diabetes 2004;53:645–53.
[36] Umpierrez GE, Casals MM, Gebhart SP, et al. Diabetic
ketoacidosis in obese African-Americans. Diabetes 1995;44:
790–5.
[37] Nyenwe E, Loganathan R, Blum S, et al. Admissions for
diabetic ketoacidosis in ethnic minority groups in a city
hospital. Metabolism 2007;56:172–8.
[38] Banerji MA, Chaiken RL, Huey H, et al. GAD antibody
negative NIDDM in adult black subjects with diabetic
ketoacidosis and increased frequency of human leukocyte
antigen DR3 and DR4. Flatbush diabetes. Diabetes 1994;43:
741–5.
 M E TAB O LI S M CL IN I CA L A N D EX PE RI ME N TA L 6 5 ( 2 01 6 ) 5 0 7–5 21
[39] Umpierrez GE, Woo W, Hagopian WA, et al. Immunogenetic
analysis suggests different pathogenesis for obese and lean
African-Americans with diabetic ketoacidosis. Diabetes
Care 1999;22:1517–23.
[40] Smiley D, Chandra P, Umpierrez GE. Update on diagno-
sis, pathogenesis and management of ketosisprone Type
2 diabetes mellitus. Diabetes Manag (Lond) 2011;1:
589–600.
[41] Kitabchi AE. Editorial: Ketosis-prone diabetes–A new sub-
group of patients with atypical type 1 an type 2 diabetes?
JCEM 2003;88(11):5087–9.
[42] Carpentier A, Mittelman SD, Lamarche B, et al. Acute
enhancement of insulin secretion by FFA in humans is lost
with prolonged FFA elevation. Am J Physiol 1999;276(6 Pt 1):
E1055–66.
[43] Bevilacqua S, Bonadonna R, Buzzigoli G, et al. Acute
elevation of free fatty acid levels leads to hepatic insulin
resistance in obese subjects. Metabolism 1987;36(5):502–6.
[44] Gosmanov AR, Smiley D, Robalino G, et al. Effects of
intravenous glucose load on insulin secretion in patients
with ketosis-prone diabetes during near-normoglycemia
remission. Diabetes Care 2010;33(4):854–60.
[45] Umpierrez GE, Smiley D, Robalino G, et al. Lack of
lipotoxicity effect on β-cell dysfunction in ketosis-prone
Type 2 diabetes. Diabetes Care 2010;33(3):626–31.
[46] Gosmanov AR, Umpierrez GE, Karabell AH, et al. Thomason
DB. Impaired expression and insulin-stimulated phosphor-
ylation of Akt-2 in muscle of obese patients with atypical
diabetes. Am J Physiol Endocrinol Metab 2004:287.
[47] Umpierrez GE, Clark WS, Steen MT. Sulfonylurea treatment
prevents recurrence of hyperglycemia in obese African–
American patients with a history of hyperglycemic crises.
Diabetes Care 1997;20(4):479–83.
[48] Banerji MA, Chaiken RL, Lebovitz HE. Prolongation of near-
normoglycemic remission in black NIDDM subjects with
chronic low-dose sulfonylurea treatment. Diabetes 1995;
4(44):466–70.
[49] Boutin P, Gresh L, Cisse A, et al. Missense mutation
Gly574Ser in the transcription factor HNF-1α is a marker of
atypical diabetes mellitus in African–American children.
Diabetologia 1999;42(3):380–1.
[50] Mauvais-Jarvis F, Smith SB, Le May C, et al. PAX4 gene
variations predispose to ketosis-prone diabetes. Hum Mol
Genet 2004;13(24):3151–9.
[51] Sobngwi E, Gautier JF, Kevorkian JP, et al. High prevalence of
glucose 6-phosphate dehydrogenase deficiency without
gene mutation suggests a novel genetic mechanism
predisposed to ketosis-prone diabetes. J Clin Endocrinol
Metab 2005;90:4446–51.
[52] Barnes AJ, Bloom SR, Goerge K, et al. Ketoacidosis in
pancreatectomized man. N Engl J Med 1977;296:1250–3.
[53] Miles JM, Rizza RA, Haymond MW, et al. Effects of acute
insulin deficiency on glucose and ketone body turnover in
man: evidence for the primacy overproduction of glucose
and ketone bodies in the genesis of diabetic ketoacidosis.
Diabetes 1980;29:926–30.
[54] Kreisburg RA. Diabetic ketoacidosis: new concepts and
trends in pathogenesis and treatment. Ann Intern Med 1978;
88:681–95.
[55] Exton JH. Mechanisms of hormonal regulation of hepatic
glucose metabolism. Diabetes Metab Rev 1987;3:163–83.
[56] Meyer C, Stumvoll M, Nadkarni V, et al. Abnormal renal and
hepatic glucose metabolism in type 2 diabetes mellitus. J
Clin Invest 1998;102:619–24.
[57] Arieff AI, Carroll HJ. Nonketotic hyperosmolar coma with
hyperglycemia: Clinical features, pathophysiology, renal
function, acid–base balance, plasma-cerebrospinal fluid
equilibria and the effects of therapy in 37 cases. Medicine
(Baltimore) 1972;51:73.
519
[58] Nurjhan N, Consoli A, Gerich J. Increased lipolysis and its
consequences on gluconeogenesis in non-insulin-depen-
dent diabetes mellitus. J Clin Invest 1992;89:169–75.
[59] McGarry JD, Woeltje KF, Kuwajima M, et al. Regulation of
ketogenesis and the renaissance of carnitine palmitoyl
transferase. Diabetes Metab Rev 1989;5:271–84.
[60] Balasse EO, Fery F. Ketone body production and disposal:
effects of fasting, diabetes, and exercise. Diabetes Metab Rev
1989;5:247–70.
[61] Moeller N, Schmitz O, Moeller J, et al. Dose–response studies
on metabolic effects of a growth hormone pulse in humans.
Metabolism 1992;41:172–5.
[62] Lu J, Zello GA, Randell E, et al. Closing the anion gap:
contribution of D-lactate to diabetic ketoacidosis. Clin Chim
Acta 2011;412:286–91.
[63] Rosival V. Interesting Development in the pathophysiology
of diabetic ketoacidosis. J Diabetes Metab 2014;5:11. http://
dx.doi.org/10.4172/2155-6156.10004155.
[64] Atchley DW, Loeb RF, Richards DW, et al. A detailed study of
electrolyte balances following withdrawal and
reestabilishment of insulin therapy. J Clin Invest 1933;12:
681–95.
[65] Howard RL, Bichet DG, RW S. Hypernatremic polyuric states.
In: D S, G G, editors. The Kidney: Physiology and patho-
physiology. New York Raven; 1991. p. 1578.
[66] DeFronzo RA, Cooke CR, Andres R, et al. The effect of insulin
on renal handling of sodium, potassium, calcium and
phosphate in man. J Clin Invest 1975;55:845–55.
[67] Kitabchi AE, Fisher JN. Insulin therapy of diabetic
ketoacidosis: physiologic versus pharmacologic doses of
insulin and their routes of administration. In: Brownlee M,
editor. Handbook of Diabetes Mellitus. New York: Garland
ATPM Press; 1981. p. 95–149.
[68] Stentz FB, Umpierrez GE, Cuervo R, et al. Proinflammatory
cytokines, markers of cardiovascular risks, oxidative stress,
and lipid peroxidation in patients with hyperglycemic
crises. Diabetes 2004;53:2079–86.
[69] Stentz FB, Kitabchi AE. Hyperglycemia-induced activation of
human T-lymphocytes with de novo emergence of insulin
receptors and generation of reactive oxygen species.
Biochem Biophys Res Commun 2005;335:491–5.
[70] Kitabchi AE, Stentz FB, Umpierrez GE. Diabetic ketoacidosis
induces in vivo activation of human T-lymphocytes.
Biochem Biophys Res Commun 2004;315:404–7.
[71] Fulop M, Rosenblatt A, Kreitzer SM, et al. Hyperosmolar
nature of diabetic coma. Diabetes 1975;24:594–9.
[72] Edge JA, Roy Y, Bergomi A, et al. Conscious level in children with
diabetic ketoacidosis is related to severity of acidosis and not to
blood glucose concentration. Pediatr Diabetes 2006;7:11–5.
[73] Rosival V. The influence of blood hydrogen ion concentra-
tion on the level of consciousness in diabetic ketoacidosis.
Ann Clin Res 1987;19:23–5.
[74] Nyenwe EA, Razavi LN, Kitabchi AE, et al. Acidosis: the
prime determinant of depressed sensorium in diabetic
ketoacidosis. Diabetes Care 2010;33:1837–9.
[75] Umpierrez G, Freire AX. Abdominal pain in patients with
hyperglycemic crises. Diabetes Care 2004;27:1873–8.
[76] Campbell IW, Duncan LJ, Innes JA, et al. Abdominal pain in
diabetic metabolic decompensation. Clinical significance.
JAMA 1975;233:166–8.
[77] SA S. Letter to the editor: a new range for the anion gap. Ann
Intern Med 1995;123:807 [New range for the anion gap].
[78] Kitabchi AE, Umpierrez GE, Miles JM, et al. Hyperglycemic
crises in adult patients with diabetes. Diabetes Care 2009;
32(7):1335–43.
[79] Brandenburg MA, Dire DJ. Comparison of arterial and
venous blood gas values in the initial emergency depart-
ment evaluation of patients with diabetic ketoacidosis. Ann
Emerg Med 1998;31:459–65.
520 M ET AB O LI S M CL IN I CA L A N D E XP E RI ME N TAL 6 5 ( 2 01 6 ) 5 0 7–52 1
[80] Kelly AM. The case for venous rather than arterial blood
gases in diabetic ketoacidosis. Emerg Med Australas 2006;18:
64–7.
[81] Nadler OA, Kinkelstein MJ, Reid SR. How well does serum
bicarbonate concentration predict the venous pH in chil-
dren being evaluated for diabetic ketoacidosis. Pediatr
Emerg Care 2011;27:907–10.
[82] Kelly AM, McAlpine R, Kyle E. Agreement between bicar-
bonate measured on arterial and venous blood gases. Emerg
Med Australas 2004;6:407–9.
[83] Nyenwe EA, Wan JY, Kitabchi AE. Venous serum bicarbon-
ate concentration predicts arterial PH in adults with diabetic
ketoacidosis. Endocr Pract 2014;20(3):201–6.
[84] Soleimanpour H, Taghizadieh A, Niafar M, et al. Predictive
value of capnography for suspected diabetic ketoacidosis in
the emergency department. West J Emerg Med 2013;14(6):
590–4.
[85] Kitabchi AE, Umpierrez GE, Murphy MB, et al. Management
of Hyperglycemic Crises in Patients with Diabetes. Diabetes
Care 2001;24:131–53.
[86] Slovis CM, Mark VG, Slovis RJ, et al. Diabetic ketoacidosis &
infection leukocyte count and differential as early predic-
tors of infection. Am J Emerg Med 1987;5:1–5.
[87] Kaminska ES, Pourmoabbed G. Spurious laboratory values in
diabetic ketoacidosis and hyperlipidaemia. Am J Emerg Med
1993;11:77–80.
[88] Rumbak MJ, Hughes TA, Kitabchi AE.
Pseudonormoglycaemia in diabetic ketoacidosis with ele-
vated triglycerides. Am J Emerg Med 1991;9:61–3.
[89] DeFronzo RA, Matzuda M, Barret E. Diabetic Ketoacidosis: A
Combined Metabolic-Nephrologic Approach to Therapy.
Diabetes Rev 1994;2:209–38.
[90] Adrogue HJ, Lederer ED, Suki WN, et al. Determinants of
plasma potassium levels in diabetic ketoacidosis. Medicine
1986;65:163–71.
[91] Atchley DW, Loeb RF, Richards DW, et al. A detailed study of
electrolyte balance following withdrawal and reestablish-
ment of insulin therapy. J Clin Invest 1933;12:297–321.
[92] Ssadi FK, John EG, Fornell L, et al. Falsely elevated serum
creatinine concentration in ketoacidosis. J Pediatr 1985;107:
562–4.
[93] Gerard SK, Khayam-Bashi H. Characterization of creatinine
error in ketotic patients: a prospective comparison of
alkaline picrate methods with an enzymatic method. Am J
Clin Pathol 1985;84:659–61.
[94] Yadav D, Nair S, Norkus EP. Nonspecific hyperamylasemia in
diabetic ketoacidosis: incidence and correlation with biochem-
ical abnormalities. Am Coll Gastroenterol 2000;95:3123.
[95] Vinicor F, Lehrner LM, Karn RC, et al. Hyperamylasemia in
diabetic ketoacidosis: sources and significance. Ann Intern
Med 1979;91:200–4.
[96] Sheikh-Ali M, Karon BS, Basu A, et al. Can serum beta-
hydroxybutyrate be used to diagnose diabetic ketoacidosis?
Diabetes Care 2008;31:643–7.
[97] Csako G, Elin RJ. Unrecognized false-positive ketones from
drugs containing free sulfhydryl groups. JAMA 1993;269:1634.
[98] Kitabchi AE, Fisher JN, Murphy MB, Rumbak MJ. Diabetic
Ketoacidosis and the Hyperglycemic Hyperosmolar
Nonketotic State. In: Kahn CR, Weir GC, editors. Joslin's
Diabetes Mellitus. 13th ed. Philadelphia: Lea & Febiger; 1994.
p. 738–70.
[99] Bjellerup P, Kaliner A, Kollind M. GLC Determination of
Serum-Ethylene Glycol Interferences in Ketotic Patients. J
Toxicol Clin Toxicol 1994;32:85–7.
[100] Kitabchi AE, Murphy MB, Spencer J, et al. Is a priming dose of
insulin necessary in a low dose insulin protocol for the
treatment of diabetic ketoacidosis? Diabetes Care 2008;31:
2081–5.
[101] Hillman K. Fluid Resuscitation in Diabetic Emergencies–A
Reappraisal. Intensive Care Med 1987;13:4–8.
[102] Martin HE, Smith K, Wilson ML. The fluid and electrolyte
therapy of severe diabetic acidosis and ketosis; a study of
twenty-nine episodes (twenty-six patients). Am J Med 1958;
24:376–89.
[103] Perel P, Roberts I. Colloids versus crystalloids for fluid
resuscitation in critically ill patients. Cochrane Database
Syst Rev 2007;17:CD000567.
[104] Caputo DG, Villarejo F, Valle GB, et al. Hydration in diabetic
ketoacidosis. What is the effect of the infusion rate?
Medicina (B Aires) 1997;57:15–20.
[105] Krentz AJ, Hale PJ, Singh BM, et al. The effect of glucose and
insulin infusion on the fall of ketone bodies during
treatment of diabetic ketoacidosis. Diabet Med 1989;6:31–6.
[106] Bratusch-Marrain PR, Komajati M, Waldhausal W. The Effect
of Hyperosmolarity on Glucose Metabolism. Pract Cardiol
1985;11:153–63.
[107] Waldhausl W, Kleinberger G, Korn A, et al. Severe Hyperglyce-
mia: Effects of Rehydration on Endocrine Derangements and
Blood Glucose Concentration. Diabetes 1979;28:577–84.
[108] Kitabchi AE, Umpierrez GE, Fisher JN, et al. Thirty years of
personal experience in hyperglycemic crises: diabetic
ketoacidosis and hyperglycemic hyperosmolar state. J Clin
Endocrinol Metab 2008;93(5):1541–52.
[109] Alberti KG, Hockaday TD, Turner RC. Small doses of
intramuscular insulin in the treatment of diabetic “coma”.
Lancet 1973;II:515–22.
[110] Kitabchi AE, Ayyagari V, Guerra SMO. Medical House Staff.
The efficacy of low dose versus conventional therapy of
Insulin for treatment of diabetic ketoacidosis. Ann Int Med
1976;84:633–8.
[111] Umpierrez GE, Latif K, Stoever J, et al. Efficacy of subcuta-
neous insulin lispro versus continuous intravenous regular
insulin for the treatment of patients with diabetic
ketoacidosis. Am J Med 2004;17:291–6.
[112] Umpierrez GE, Cuervo R, Karabell A, et al. Treatment of
diabetic ketoacidosis with subcutaneous insulin aspart.
Diabetes Care 2004;27:1873–8.
[113] Della Manna T, Steinmetz L, Campos PR, et al. Subcutaneous
use of a fast-acting insulin analog: an alternative treatment
for pediatric patients with diabetic ketoacidosis. Diabetes
Care 2005;28:1856–61.
[114] Umpierrez GE, Jones S, Smiley D, et al. Insulin analogs
versus human insulin in the treatment of patients with
diabetic ketoacidosis: a randomized controlled trial. Diabe-
tes Care 2009;32:1164–9.
[115] Ersöz HO, Ukinc K, Köse M, et al. Subcutaneous lispro and
intravenous regular insulin treatments are equally effec-
tive and safe for the treatment of mild and moderate
diabetic ketoacidosis in adult patients. Int J Clin Pract 2006;
60:429–33.
[116] Savage MW, Dhatariya KK, Kilvert A, et al. Joint British
Diabetes Societies guideline for the management of diabetic
ketoacidosis. Diabet Med 2011;28:508–15.
[117] Doshi P, Potter AJ, De Los SD, et al. Prospective randomized
trial of insulin glargine in acute management of diabetic
ketoacidosis in the emergency department: a pilot study.
Acad Emerg Med 2015;6:657–62.
[118] Wiggan MI, O'Kane MJ, Harper R, et al. Treatment of diabetic
ketoacidosis using normalization of blood 3
hydroxybutyrate concentration as the endpoint of emer-
gency management. Diabetes Care 1997:1347–52.
[119] Adrogue HJ, Lederer ED, Suki WN, et al. Determinants of
plasma potassium levels in diabetic ketoacidosis. Medicine
1986;65:163–71.
[120] Beigelman PM. Potassium in severe diabetic ketoacidosis
(editorial). Am J Med 1973;54:419–20.
 M E TAB O LI S M CL IN I CA L A N D EX PE RI ME N TA L 6 5 ( 2 01 6 ) 5 0 7–5 21
[121] Morris LR, Murphy MB, Kitabchi AE. Bicarbonate therapy in
severe diabetic ketoacidosis. Ann Intern Med 1986;105:
836–40.
[122] Viallon A, Zeni F, Lafond P, et al. Does bicarbonate therapy
improve the management of severe diabetic ketoacidosis?
Crit Care Med 1999;27:2690–3.
[123] Hale PJ, Crase J, Nattrass M. Metabolic effects of bicarbonate
in the treatment of diabetic ketoacidosis. Br Med J (Clin Res
Ed) 1984;289:1035–8.
[124] Okuda Y, Adrogue HJ, Field JB, et al. Counterproductive
effects of sodium bicarbonate in diabetic ketoacidosis in
childhood. J Clin Endocrinol Metab 1996;81:314.
[125] Chua HR, Schneider A, Bellomo R. Bicarbonate in diabetic
ketoacidosis - a systematic review. Ann Intensive Care 2011;
6(1(1)):23.
[126] Fisher JN, Kitabchi AE. A randomized study of phosphate
therapy in the treatment of diabetic ketoacidosis. J Clin
Endocrinol Metabol 1983;57:177–80.
[127] Wilson HK, Keuer SP, Lea AS, et al. Phosphate therapy in
diabetic ketoacidosis. Arch Intern Med 1982;142:517–20.
[128] Gibby OM, Veale KEA, Hayes TM, et al. Oxygen availability
from the blood and the effect of phosphate replacement on
erythrocyte 2–3 diphosphoglycerate and hemoglobin-oxy-
gen affinity in diabetic ketoacidosis. Diabetologia 1978;15:
381.
[129] Miller DW, Slovis CM. Hypophosphatemia in the emergency
department therapeutics. Am J Emerg Med 2000;18:457–61.
[130] Oh MS, Carroll HJ, Goldstein DA, et al. Hyperchloremic
acidosis during the recovery phase of diabetic ketosis. Ann
Intern Med 1978;89(6):925–7.
[131] Oh MS, Carroll HJ, Uribarri J. Mechanism of normochloremic
and hyperchloremic acidosis in diabetic ketoacidosis.
Nephron 1990;54:1–6.
[132] Glaser N, Barnett P, McCaslin I, et al. Risk factors for cerebral
edema in children with diabetic ketoacidosis. The Pediatric
Emergency Medicine Collaborative Research Committee of the
American Academy of Pediatrics. N Engl J Med 2001;344:264–9.
[133] Silver SM, Clark EC, Schroeder BM, et al. Pathogenesis of
cerebral edema after treatment of diabetic ketoacidosis.
Kidney Int 1997;51:1237–44.
[134] Glaser NS, Wooten-Gorges SL, Marcin JP, et al. Mechanism of
cerebral edema in children with diabetic ketoacidosis. J
Pediatr 2004;145(2):149–50.
521
[135] Glaser NS, Wooten-Gorges SL, Buonocore MH, et al. Fre-
quency of subclinical cerebral edema in children with
diabetic ketoacidosis. Pediatr Diabetes 2006;7(2):75–80.
[136] Rose KL, Watson AJ, Drysdale TA, et al. Simulated diabetic
ketoacidosis therapy in vitro elicits brain cell swelling via
sodium-hydrogen exchange and anion transport. Am J
Physiol Endocrinol Metab 2015;309(4):E370–9.
[137] Isales CM, Min L, Hoffman WH. Acetoacetateand
betahydroxybutyrate differentially regulate endothelin-1
and vascular endothelial growth factor in mouse brain
microvascular endothelial cells. J Diabet Complications
1993;13:91.
[138] Büyükaşik Y, Ileri NS, Haznedaroğlu IC, et al. Enhanced
subclinical coagulation activation during diabetic
ketoacidosis. Diabetes Care 1998;21:868–70.
[139] Gibb FW, Teoh W, Graham J, et al. Previous Diabetic
Ketoacidosis Is Associated with a Significant Risk of Death
in the following 5 Years [Abstract]. Diabetes 2015;64(Suppl 1
273R).
[140] Zinman B, Wanner C, Lachin JM, et al. Empagliflozin,
Cardiovascular Outcomes, and Mortality in Type 2 Diabetes.
N Engl J Med 2015 [Epub ahead of print].
[141] Musey VC, Lee JK, Crawford R, et al. Diabetes in urban
African-Americans. Cessation of insulin therapy is the
major precipitating cause of diabetic ketoacidosis. Diabetes
Care 1995;18:483–9.
[142] Maldonado MR, Chong ER, Oehl MA, et al. Economic Impact
of Diabetic Ketoacidosis in a Multiethnic Indigent Popula-
tion: Analysis of costs based on the precipitating cause.
Diabetes Care 2003;26:1265–9.
[143] Laffel LM, Brackett J, Ho J, et al. Changing The Process of
Diabetes Care Improves Metabolic Outcomes and Reduces
Hospitalizations. Qual Manag Health Care 1998;6:53–62.
[144] Ellis D, Naar-King S, Templin T, et al. Multisystemic therapy
for adolescents with poorly controlled type 1 diabetes:
reduced diabetic ketoacidosis admissions and related costs
over 24 months. Diabetes Care 2008;31:1746–7.
[145] Warner GA, Greene GS, Buschsbaum MS, et al. (1998).
Diabetic ketoacidosis associated with cocaine use. Arch
Intern Med 1998;158:1799–802.
[146] Nyenwe EA, Kitabchi AE. Evidence-based management of
hyperglycemic emergencies in diabetes mellitus. Diabetes
Res Clin Pract 2011;94:340–51.